Research On Flow Field Analysis And Endocrine Fault Diagnosis Of Damaged Hydraulic Cylinder Seal Ring

The hydraulic cylinder is an execution component widely used in the hydraulic system of construction machinery.When it fails,it will affect the working efficiency of the construction machinery,even cause a major accident.The internal leakage caused by the failure of the seal of the hydraulic cylinder is the most common failure mode of the hydraulic cylinder.So the research on fault simulation and fault diagnosis of hydraulic cylinder is of great significance.The main research contents of this paper are as follows:1.Analysing the factors that cause damage to the hydraulic cylinder seal;Taking ZK-PH hydraulic cylinder as the research object,artificially prefabricating the damage of UHS type seal.The hydraulic cylinder leakage test was carried out based on the RCYCS-B test bench.The fluid model was established according to the prefabricated seal damage failure in the experiment.In the Workbench simulation software,the fluid simulation module Fluent was used to analyze the flow field of the seal damage fluid model and calculate the internal leakage.Comparing the experimental measurements with the calculated internal leakage,the results show that:In the physical experiment and Fluent simulation,the internal leakage amount of the hydraulic cylinder seal is consistent with the increase of the hydraulic system pressure.It is proved that the method of using Fluent to simulate seal damage and internal leakage is feasible.2.Taking the DAS combination seal ring used in the QY110 truck crane leg cylinder as the research object,three kinds of damage failure modes were artificially prefabricated.Based on the pressure sensor,data acquisition card,signal conditioning module and Labview software in the working condition simulation test bench,the pressure signals of the rodless cavity of the hydraulic cylinder under various fault conditions are collected.Based on the principle of wavelet transform,making four kinds of eigenvalues which are extracted from non-stationary pressure signals:wavelet packet energy,wavelet packet energy entropy,wavelet packet energy variance and wavelet coefficient d₄ rms value,as the feature vector of a fault diagnosis for seal damage;According to the experimental three prefabricated three seal damage faults,the fluid model is established.The simulation module Fluent is used to simulate the fluid under the fault condition of the seal ring,and collecting the pressure signal of the rodless cavity in the fluid model,getting five kinds of eigenvalues which are extracted from non-stationary pressure signals:wavelet packet energy,wavelet packet energy entropy,wavelet packet energy variance and wavelet coefficient d₁ and d₂ rms value.Providing the basis for the next failure pattern recognition.3.Combining endocrine with wavelet neural network and using endocrine strategy to adjust weights,a new endocrine neural network algorithm is proposed and applied to the fault diagnosis of seal damage,and the recognition effect of support vector machine is compared.The results show that the two algorithms achieve the same diagnostic accuracy when troubleshooting the damage state of the seal in the physical experiment.When the fault is detected in the simulation,the fault identification accuracy of the endocrine neural network is higher than the support vector machine,which verifies the effectiveness of endocrine neural network algorithm in hydraulic cylinder fault diagnosis.